low cost 1kw mains grid tie inverter

[aat39]

Hey guys, im currently studying electrical engineering in New Zealand. I am trying to design a 1kW , 48 volts DC to 230 volts AC grid tied inveter that does not use a battery bank. We need short circuit and no load fault protection. The input power can range from 100 watts to 1200 watts max as will be tapped off a renewable energy source (will come online at 100 watts). Have been looking at using DC-DC amplification of the voltage source then using IGBT's in SPWM to convert to ac mains using a controller based off the mains itself to get synchronisation.

My question is what is the best DC-DC converter style to use for this, we are currently looking at flyback converters but are having problems with modelling them using TINA.

Have read some of the posts relating to this topic but I seem to be going around in circles with no definitive solution.

Is the flyback converter the best/cheapest to use for this purpose? am using DC-DC then DC-AC rather than DC-AC AC-AC as am lead to believe a huge cost is in the transformer where with high switching frequencies of the flyback the transformer can be kept small and therefore offer a cheaper device. Thanks for your help in advance.

 

[Mr RB]

Flyback converter is not a good choice for any decent power transfer of low voltage to high voltage. Try a push-pull DC-DC converter.

Please post any schematics/pictures, there are a few people here interested in grid tie inverter systems.

 

[aat39]

Thanks for the reply, with these dc converters whats the limiting factor on power throughput. Found that flyback only goes up to about 150watt where the push pull goes from 100-1000watts (thanks wikipedia) . Also 1200 watts will be fine to put through a push pull schematic if the correct components are used?

 

[tcmtech]

Is the flyback converter the best/cheapest to use for this purpose? am using DC-DC then DC-AC rather than DC-AC AC-AC as am lead to believe a huge cost is in the transformer where with high switching frequencies of the flyback the transformer can be kept small and therefore offer a cheaper device. Thanks for your help in advance.

I am not sure how exactly they control it in the direct line connect systems where all of the AC synchronization is done on the line side of the step up system. As far as I know there still has to be a fair amount of magnetics based filtering on the AC line regardless.

Given the more basic and simple approach I have been using in my designs they need that big transformer to work as a line filter and waveform shaping device in order to keep line harmonics and noise below threshold levels. For me there is no point in doing away with the component that does the voltage matching, all of the primary cleanup work, and gets the output to within acceptable standards and does it all in one package with no complex circuitry or control components.
Plus a good quality iron core transformer is still about the same size and power capacity of its solid state equivalent and has equal and often time greater average efficiency as well. Its also far more robust, works in step up and step down operation, inherently likes to make a sine wave with only one additional capacitor, and has considerable short term overload capacities as well.

And as far as expensive that relative. I dont find a 1KVA transformer the least bit expensive at around $80 - $130 new. Considering the rest of the circuitry costs about that much as well. And a new factory GTI unit is still pushing near $1000 a KW.

But over all it does come down to what each person wants to do. For me simple functional and robust is what I prefer. Its proven to me it works well and reliably plus has the least overall parts count in order to do the intended functions.

I am still interested to see any other versions of a working home made GTI system though!

 

[Warpspeed]

My question is what is the best DC-DC converter style to use for this, we are currently looking at flyback converters but are having problems with modelling them using TINA.

It depends...............

If you are running fairly high power all the time, push pull dc/dc square wave inverters will have much smaller magnetics, and are the best choice for size and cost.
But at much reduced power levels they can be horribly inefficient. The problem is the magnetising current.
Even with zero output load current, it may take significant input current to swing the transformer back and forth over it's peak to peak voltage range. So a 1Kw inverter may draw 100 watts even with zero load. That may concern you, or it may not.

Now flybacks are very interesting, and especially discontinuous flybacks.
They can work over very large ranges of both input and output voltage, and they are extremely efficient at very low power levels. The only problem is the size of the magnetics required at high power levels.
But you can overcome that by using several smaller flyback operated in parallel.

Two flybacks can be run 180 degrees out of phase, to reduce the ripple current at both input and output. With two, you get twice the ripple frequency at half the amplitude, which is MUCH easier to filter with capacitors. You can run four 90 degrees out of phase, or any number.

It also offers some redundancy. Four 250 watt dc/dc flyback modules would produce 1Kw of dc with low ripple at both input and output. And at extremely low power you could completely turn off modules, thus increasing the low power efficiency even more.

Flybacks operate in current mode, and can be designed to be very reliable. Push pull square wave inverters can generate some horrific current spikes in the switching devices, especially if the transformer saturates. Flybacks are much more benign.